Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Update docs #1806

Merged
merged 8 commits into from Jan 3, 2017

Conversation

Projects
None yet
5 participants
@ajdapretnar
Copy link
Contributor

commented Dec 8, 2016

Issue

Some widgets missing documentation.

Description of changes

Documentation for Transpose, SQL Table, Manifold Learning, and Univariate Polynomial Regression. Classification Tree moved to proper place.

Includes
  • Code changes
  • Tests
  • Documentation
@ajdapretnar

This comment has been minimized.

Copy link
Contributor Author

commented Dec 8, 2016

Should we remove Univariate Polynomial Regression altogether, since it's a worse duplicate of Orange3-Educational/Polynomial_Regression?

@janezd

This comment has been minimized.

Copy link
Contributor

commented Dec 8, 2016

Dear @ajdapretnar, I must strongly object to the language used in your previous comment. "Univariate Polynomial Regression" is not a worse duplicate of "Educational/Polynomial Regression". Quite twice the opposite. "Educational/Polynomial Regression" is a better duplicate "Univariate Polynomial Regression".

Please stop spreading fake fake news before somebody starts shoot shooting in the lab - which, as you well know, is quite likely to happen.

(But to answer your question: yes, sure, that was the intention.)

@ajdapretnar

This comment has been minimized.

Copy link
Contributor Author

commented Dec 9, 2016

@janezd I apologize for my blatant assumptions. Grasshopper is to learn. I actually meant to suggest UniPolReg is a proud father of Edu/PolyReg. 🎓

(So I wait?)

@janezd

This comment has been minimized.

Copy link
Contributor

commented Dec 9, 2016

Apology and explanation accepted.

(Now that we made this clear, you can remove the father widget.)

@ajdapretnar

This comment has been minimized.

Copy link
Contributor Author

commented Dec 9, 2016

As for removal, I can surely do it in docs, but how about the widget itself? As far as I know it's not just a simple delete. We need to map to Edu/PolyReg, right?

@kernc

This comment has been minimized.

Copy link
Member

commented Dec 9, 2016

This PR adds 1.3 MB of binary change.

@ajdapretnar

This comment has been minimized.

Copy link
Contributor Author

commented Dec 9, 2016

I still don't know what to do about it. And these are not even old files, they're all new.

@kernc

This comment has been minimized.

Copy link
Member

commented Dec 9, 2016

@kernc

This comment has been minimized.

Copy link
Member

commented Dec 9, 2016

The PR contains map.png and map-stamped.png but no map.rst?

@ajdapretnar

This comment has been minimized.

Copy link
Contributor Author

commented Dec 9, 2016

Yeah, map.rst go lost somewhere in space-time to the anger of its author, who spent a great deal of time working on it. Here we go again....

@ajdapretnar ajdapretnar force-pushed the ajdapretnar:update-docs branch from 7b6d019 to 37715bd Dec 12, 2016

@ajdapretnar ajdapretnar changed the title [WIP] Update docs Update docs Dec 12, 2016

@ajdapretnar

This comment has been minimized.

Copy link
Contributor Author

commented Dec 12, 2016

This is done, I guess. I have indexed newly added images. Would it make sense to do old ones as well?

Please review and merge.

@codecov-io

This comment has been minimized.

Copy link

commented Dec 12, 2016

Current coverage is 89.20% (diff: 100%)

Merging #1806 into master will not change coverage

@@             master      #1806   diff @@
==========================================
  Files            86         86          
  Lines          9077       9077          
  Methods           0          0          
  Messages          0          0          
  Branches          0          0          
==========================================
  Hits           8097       8097          
  Misses          980        980          
  Partials          0          0          

Sunburst

Powered by Codecov. Last update 48e7e1c...b44c46b

.. figure:: images/map-stamped.png

1. Define map properties:
- Set the type of *map*: Black and White, `OpenStreetMap <http://www.openstreetmap.org/#map=5/51.500/-0.100>`_, Topographic, Satellite, Print, Light, Dark, Railyways and Watercolor.

This comment has been minimized.

Copy link
@kernc

kernc Dec 12, 2016

Member

What's with the OSM link? If you do include it, perhaps make it position neutral (truncate the URL fragment)?


1. Define map properties:
- Set the type of *map*: Black and White, `OpenStreetMap <http://www.openstreetmap.org/#map=5/51.500/-0.100>`_, Topographic, Satellite, Print, Light, Dark, Railyways and Watercolor.
- Set latitude and longitude attributes, if the widget didn't recognize them automatically.

This comment has been minimized.

Copy link
@kernc

kernc Dec 12, 2016

Member

Mention somewhere that lats are -90 (S) to 90 (N) and longs -180 (W) to 180 (E). We've come upon datasets with other origins floating about.

- Set the type of *map*: Black and White, `OpenStreetMap <http://www.openstreetmap.org/#map=5/51.500/-0.100>`_, Topographic, Satellite, Print, Light, Dark, Railyways and Watercolor.
- Set latitude and longitude attributes, if the widget didn't recognize them automatically.
2. Overlay:
- Set the target (class) for predictive mapping. A learner has to be provided on the input.

This comment has been minimized.

Copy link
@kernc

kernc Dec 12, 2016

Member

Perhaps mention that the learner is trained on lats and longs only, i.e. it maps latlon pairs to this selected attribute.

- Opacity: set transparency of data points
- Symbol size: size of data points (small to large)
- Jittering: disperse overlaid data points
- Cluster points: cluster neighboring points with `naive greedy clustering <https://github.com/Leaflet/Leaflet.markercluster>`_ (available when zoomed in)

This comment has been minimized.

Copy link
@kernc

kernc Dec 12, 2016

Member

available when zoomed in

More precisely, when less than 600 points are in the view.


In the first example we will model class predictions on a map. We will use *philadelphia-crime* data set, load it with :doc:`File <../data/file>` widget and connect it to **Map**. We can already observe the mapped points in Map. Now, we connect :doc:`Classification Tree <../classify/classificationtree>` to Map and set target variable to Type. This will display the predicted type of crime for a specific region of Philadelphia city (each region will be colored with a corresponding color code, explained in a legend on the right).

.. figure:: images/map-classification.png

This comment has been minimized.

Copy link
@kernc

kernc Dec 12, 2016

Member

I'd color the points as well.

@kernc

This comment has been minimized.

Copy link
Member

commented Dec 12, 2016

This now adds 2.2 MB to repository size. The images do not appear to be indexed.

% ll collage-manifold*
-rw-r--r-- 1 ku ku  70K Dec 12 16:08 collage-manifold-indexed.png
-rw-r--r-- 1 ku ku 271K Dec 12 16:08 collage-manifold.png
@ajdapretnar

This comment has been minimized.

Copy link
Contributor Author

commented Dec 12, 2016

Ah, this on isn't, missed it, the rest should be.

@kernc

This comment has been minimized.

Copy link
Member

commented Dec 12, 2016

You sure only that one?

% ll map-classification*
-rw-r--r-- 1 ku ku 127K Dec 12 16:35 map-classification-indexed.png
-rw-r--r-- 1 ku ku 520K Dec 12 16:35 map-classification.png
% ll tree-viewer-classification*
-rw-r--r-- 1 ku ku 27K Dec 12 16:36 tree-viewer-classification-indexed.png
-rw-r--r-- 1 ku ku 70K Dec 12 16:36 tree-viewer-classification.png

Manual methods don't work long term.


Would you like to try code a CI test that checks docs images indeed have indexed palettes?

@ajdapretnar

This comment has been minimized.

Copy link
Contributor Author

commented Dec 12, 2016

Would you like to try code a CI test that checks docs images indeed have indexed palettes?

Um, ok?
I guess I was using GIMP wrong. Overwrite 'file.png' doesn't actually overwrite the file. #intuitive

@ajdapretnar ajdapretnar force-pushed the ajdapretnar:update-docs branch from 37715bd to 7958c35 Dec 14, 2016

@ajdapretnar

This comment has been minimized.

Copy link
Contributor Author

commented Dec 14, 2016

How about now?

@janezd janezd self-assigned this Dec 23, 2016

@kernc

This comment has been minimized.

Copy link
Member

commented Dec 23, 2016


**Outputs**:

- **Transformed Data**
- **Transformend Data**

This comment has been minimized.

Copy link
@janezd

janezd Dec 23, 2016

Contributor

n

Examples
--------

Below, is a simple schema, where we have read the data, constructed the

This comment has been minimized.

Copy link
@janezd

janezd Dec 23, 2016

Contributor

Is the first comma necessary?

This comment has been minimized.

Copy link
@ajdapretnar

ajdapretnar Dec 28, 2016

Author Contributor

I suppose so. This was done by @karoema, so it's a professional review. I just moved the file to appropriate place. :)


Below, is a simple schema, where we have read the data, constructed the
classification tree and viewed it in our tree viewer. If both the viewer
and :doc:`Classification Tree <../classify/classificationtree>` are open, any run of the tree induction

This comment has been minimized.

Copy link
@janezd

janezd Dec 23, 2016

Contributor

Perhaps re-run instead of run?

@ajdapretnar ajdapretnar force-pushed the ajdapretnar:update-docs branch from 7958c35 to b44c46b Dec 28, 2016

@ajdapretnar

This comment has been minimized.

Copy link
Contributor Author

commented Dec 28, 2016

Fixed and rebased.

@ajdapretnar ajdapretnar referenced this pull request Dec 28, 2016

Merged

[ENH] Map widget #1735

2 of 3 tasks complete
@kernc

kernc approved these changes Dec 28, 2016

@lanzagar lanzagar merged commit d96c00e into biolab:master Jan 3, 2017

5 checks passed

codecov/patch Coverage not affected when comparing 48e7e1c...b44c46b
Details
codecov/project 89.20% (+0.00%) compared to 48e7e1c
Details
continuous-integration/appveyor/pr AppVeyor build succeeded
Details
continuous-integration/travis-ci/pr The Travis CI build passed
Details
licence/cla Contributor License Agreement is signed.
Details

@ajdapretnar ajdapretnar deleted the ajdapretnar:update-docs branch Feb 20, 2017

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
You can’t perform that action at this time.